Int. J. Adv. Sci. Eng. Vol.8 No.4 2425-2430 (2022) 2425 E-ISSN: 2349 5359; P-ISSN: 2454-9967
Avian Frank & Ramesh
International Journal of Advanced Science and Engineering www.mahendrapublications.com
Design and Analysis of Multiplier Accumulator Circuit for DSP
Applications
Avian Frank, K. B. Ramesh
Department of Electronics and Instrumentation Engineering,
RV College of Engineering, Bengaluru, India.
1. INTRODUCTION
A Multiplier Accumulator (MAC) unit
performs the operations of multiplying two
numbers and accumulates the result in a
register repeatedly to perform continuous and
complex operations. MAC can speed up the
process of computation. It has numerous
applications in digital signal processing,
including filtering, and convolution. MAC also
has tremendous applications in audio and video
signal processing, Artificial Intelligence (AI),
machine learning, military and defense [1].
Since these operations require a cyclic
application of multiplication and addition, the
speed of execution depends on the overall
performance of the MAC unit [2]. Using a MAC
unit improves the accuracy and also reduces
time delay for computing dot product, matrix
multiplication, artificial neural networks, and
various mathematical computations.
The demand for fast and portable electronic
devices has been rising, as they help in
accomplishing our day to day tasks. The
computation speed of processor is highly
dependent on these arithmetic units. The
arithmetic unit complexity increases with
improvement of the speed and power
performance of the processor. Therefore, it is
essential to design the unit to reduce the
complexity based on the algorithm and number
of components used.
MAC being the fundamental unit of DSP, it
significantly affects the performance of the
system. MAC unit comprises of a multiplier and
adder. The design parameters of any
architecture completely depend on the basic
building blocks which are the multiplier and the
adder. The enhancement of these building
blocks can improve the performance of the
overall unit. Along these lines, the enhancement
of the multiplier speed and area is a noteworthy
test for the framework architects. This test can
be effectively overwhelmed by the utilization of
various multiplier techniques and appropriate
adder circuit. The objective of this paper is to
design a MAC unit to perform computation with
optimum speed, low power dissipation and chip
area.
The remainder of this paper is organized as
follows. Section 2 explain the basic operation of
MAC unit and illustrate an application of MAC.
Section 3 explains the design and performance
requirements of MAC unit. Section 4 presents
various adder circuits. Section 5 presents
various multiplier circuits.
Rakesh and Sunitha have implemented the
novel design of 32-bit MAC unit with 32-bit
ABSTRACT: Multipliers and Accumulators are an imperative component of DSP application systems. It
plays a vital role in high speed digital signal processing (DSP), image processing multiplier and
processing fast Fourier transform (FFT). These computations require large number of multiplication
and addition operation which requires dedicated MAC and Arithmetic and Logic Unit (ALU)
architectures. Multipliers and adders are the key components of these arithmetic units as it determines
the overall performance of the system, i.e. speed, power and area consumed. This paper reviews the
components of MAC circuits analyzing and comparing the performance parameters in terms of area,
speed and power consumption. Carry Look Ahead adder is the most suitable for low power and for
high performance, whereas Wallace tree multiplier can be implemented for chips having larger area
and require high speed. Various other multiplier and adder circuits are also reviewed.
KEYWORDS: Multipliers and Accumulators, multiplier algorithms, adder circuits, ALU
https://doi.org/10.29294/IJASE.8.4.2022.2425-2430 ©2021 Mahendrapublications.com, All rights reserved
*Corresponding Author:avianfrank08@gmail.com
Received: 05.03.2022 Accepted: 27.04.2022 Published on: 18.05.2022